Kee Hinckley wrote:
At 1:59 PM -0500 3/6/03, Jim Youll wrote:
But if you had a proper client that hid all the work, you could "give"
a different e-mail address to every correspondent, and if it leaked
out, you need only cancel that one and give that particular
correspondent a freshly-generated address, no?
Read the thread on striker's spam problem. Hundreds of thousands of
messages a day, all bouncing due to a dictionary attack gone wrong.
We think.
Now consider what happens if everyone has hundreds of temporary
addresses that can get into spammers hands. Sure, you cancel it, but
that doesn't mean the spammer stops sending. You've made your life more
complicated, and you've made your ISPs life hell.
Yeah - to reiterate a comment that may have been lost: even tho Allan is
on "high speed", he can no longer afford to even just reject connections
at the TCP/IP layer. IIRC, he quoted something like 330Mb of NACKs per
_day_. Yow.
We can afford to handle a spamtrap like that because we're on a mondo
humunguous pipe. But if the cron job deleting the email stalls for more
than an hour, the 9Gb disk fills up and the machine falls over... I
can't even imagine AOL's 3 orders of magnitude larger spam catch must be
like... ~5Tb of data per hour. That's more than the total US backbone
flow of only a few years ago.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg